J. Am. Chem. SOC.1985, 107, 7781-7782
$5
7'0
7'5imln)
7781
345
350
3 5 5 l m t n l 360
li7O
'2'o"'ni
Figure 2. Separation of isotopic benzoic acids in recycle chromatography. Column: Cosmosil 5-CI8-P,4.6-mm i.d., 15 cm X 4. Mobile phase: 20% methanol, 0.05 M acetic acidsodium acetate buffer, pH 4.83. Flow rate: 0.8 mL/min. Temperature: 30 OC. (a) 1 cycle, (b) 5 cycles, (c) 17 cycles. Mass of samples: 0.5 fig each of 1-I6O2and 1-180,in (a) and (b); 0.25 fig of 1-160-180 and 0.5 fig each of 1-1602and 1-180, in (c). A part of the tail of the 1-I8O2peak was shaved off during the recycle operation in (c).
Figure 1. pH dependence of k'values of 1-I6O2normalized by using the k'value of benzyl alcohol in each mobile phase, and the p H dependence of I8O isotope effect on retention, I6k'/l8k'. The curves were drawn by using eq 2 by applying kLH = 2.646, k i - = 0.128, 16K, = 3.631 X and I8K, = 3.561 X which were estimated from the results between pH 2.4 and 6.3.
Equation 2 indicates that the k'of a carboxylic acid is determined by the extent of ionization of the acid. The dissociation constant of benzoic acid containing I8Ois expected to be smaller than that of l-1602.5 This will result in the lesser extent of ionization of 1-1802and 1-160-'80 compared to 1-1602at a pH near the pK,. Therefore, the retention of benzoic acids containing I8O is expected to be greater than that of 1-I6O2 in such a pH region. This assumes that the secondary oxygen isotope effects on kjAH and ktA-are small compared to the equilibrium oxygen isotope effects on acid dissociation constants K,, which influence the retention of the acid according to eq 2. The present chromatographic system consisted of an octadecylsilylated silica stationary phase6 and a methanol-water mobile phase (20/80 v/v) containing acetic acidsodium acetate buffer.' 1-I6O2and l-l8O2showed very similar k'values below pH 3 and above pH 6 (where the acids dissociate very little or dissociate almost completely). In the pH region near the pK,, however, differences of up to 1.2% in k'values were found between 1-W2 and 1-I8O2. 1-1802was actually retained longer than 1-I6O2. The maximum separation was achieved near pH 5, as shown in Figure 1. From eq 2, where ktAHand k a - were found to be similar for the two isotopic acids, eq 3 can be derived, where I6k'and I8k'
to be 1.022. Taking into account the presence of an organic solvent, the present results with benzoic acid agree well with the results for formic acid.5a Thornton also estimated the oxygen isotope effect on the acidity of H 3 0 + to be about Column switching and solute recycling were utilized to resolve the three isotopic benzoic acids at pH 4.8. Two pairs of two 15-cm columns (60-cm total) were connected through a six-port ~ a l v e . ] ~ . ~ The chromatograms after 1, 5, and 17 recycles, giving the effective column lengths of 0.6, 3.0, and 10.2 m, respectively, are shown in Figure 2. The chromatogram in Figure 2c shows the resolution of the three isotopic benzoic acids with about 340000 theoretical plates. Figure 2a shows that partial separation can be achieved without any solute recycling. The results suggest the possibility of the enrichment of minor isotopes in various dissociable compounds from the natural abundance to near 100% by the ionization control technique in RPLC.
Acknowledgment. We acknowledge Dr. J. W . Dolan for reviewing the manuscript. Registry No. C 6 H 5 C 1 6 0 1 6 0 H65-85-0; , C,H5C160180H,21 048-30-6; C6H5C180180H,17217-84-4. (9) Henry, R. A,; Byrne, S. H.; Hudson, D. R. J . Chromatogr. Sci. 1974, 12, 197-199.
Separation of Nitrogen Isotopic Compounds by Reversed-Phase Liquid Chromatography on the Basis of Nitrogen Isotope Effects on the Dissociation of Amines Nobuo Tanaka,* Atsumi Yamaguchi, and Mikio Araki Faculty of Textile Science Kyoto Institute of Technology Matsugasaki, Sakyo-ku, Kyoto 606, Japan
stand for the capacity factors of 1-I6O2and 1-I8O2,respectively. This equation yields the isotope effects, 16Ka/18Ka,based on the k'values in RPLC, k a H at low pH, k i - at high pH, and I6k'and I8k' at intermediate pH. 16Ka/18Kawas found to be 1.020 A 0.002 on the basis of the results between pH 4 and 6 in 20% methanoL8 The oxygen isotope effect on the dissociation of formic acid in water was reported (6) Cosmosil 5-CI8-P(4.6-mm i. d., 15 cm, Nakarai Chem.) was used with a Shimadzu LC-3A pump and a Waters M440 UV detector operated at 254 nm. Two 15-cm columns connected in series were used to determine k'values at various pH at 30 OC. (7) The apparent pH values of the methanol-0,0625 M acetate buffer mixtures (20/80 v/v) were measured with a pH meter. The pH values were not directly used in the calculation of the isotope effect, 16Ka/18Ka, according to eq 3. Chromatographic runs were carried out in triplicate with the reproducibility better than &0.2%in k'values. In order to obtain a reproducible pH-k' profile, all k' values were normalized using the k' value of benzyl alcohol as a standard in each mobile phase. (8) Although 1 - W 2contains two I8O atoms per molecule, only one oxygen atom is involved at a time in the dissociation process. The secondary isotope effect by the second '*Oatom in the same molecule was neglected.5a
0002-7863/85/1507-7781$01.50/0
Kazuhiro Kimata Nakarai Chemical Co., Kuze Minami- ku, Kyoto 601, Japan Received June 27, 1985 Isotope separation has been one of the most interesting subjects in separation science, and the importance of isotopic separation is increasing due to the wide use of isotopes in energy and life science oriented applications. We have reported the measurement of oxygen isotope effects on the dissociation constant of benzoic acid and the separation of the three isotopic benzoic acids, C6H,C160160H,C6H,C160180H,and C6H5C'80180Hby ionization control in reversed-phase liquid chromatography (RPLC).' We wish to report here the first example of the separation of (1) Tanaka, N.; Araki, M., J . Am. Chem. SOC.,preceding paper in this
issue.
0 1985 American Chemical Society
7782 J. Am. Chem. SOC.,Vol. 107, No. 25, 1985
Communications to the Editor I4Ka/l5Ka,in 5% methanol was estimated to be 1.019 according to eq 3 derived from eq 2,’3’ where I4k’and 15k’are the k’values
205
210
215
220fminl
840
845
850
855
860w”
Figure 1. Separation of isotopic anilines in recycle chromatography. Column: Cosmosil 5-C18-P,4.6-mm i.d., 15 cm X 4,(a) 5 cycles, (b) 20 cycles. Mobile phase: 5% methanol, 0.05 M acetate, 0.01% triethylamine, pH 4.24. Flow rate: 1 mL/min. Temperature: 30 O C . Amount of sample: 1.4 FCg each of l-I4Nand l-15N.
nitrogen isotopic compounds by RPLC and the measurement of the nitrogen isotope effect on the dissociation of aniline. Aniline-I5N (P 5 N )was completely separated from aniline-I4N (1 - 1 4 ~ on the basis of the small difference in the dissociation constants of these isotopic anilines. The nitrogen isotope effect on the dissociation of 1, I4Ka/l5Ka,was found to be 1.019 in 5% methanol. Ionization control in RPLC2 has enabled the measurement of oxygen isotope effects and the separation of oxygen isotopes based on the difference in the dissociation constants of carboxylic acids.’ A similar difference in the dissociation constants due to the difference in mass of the nitrogen atom is expected with amines. Presumably amines containing I5N are expected to be stronger bases and protonated more easily than amines containing I4N, The measurement of the nitrogen isotope effect and the separation of nitrogen isotopic compounds were attempted with l-14Nand I-l5N3by RPLC. The HPLC instrumentation employed in this study was similar to that described earlier.’ Two pairs of two 15-cm columns (60-cm total) packed with octadecylsilylated silica gel4 were used with a six-port valve in the solute recycle mode.5 Methanol-water mixtures (5/95 v/v) containing 0.02 M phosphate at pH 2.37 and 7.24 and 0.05 M acetate buffer at pH 4.24 were used as mobile phasesS6 In order to eliminate slight peak tailing associated with basic compounds on the silica-based stationary phase, 0.01% triethylamine was added to the mobile phase. The retention of amines in RPLC can be described by eq 2,* RNH3+ + H 2 0
+
RNH2 H30+ K, = [RNH,] [H3O+]/[RNH3+]
k’=
k’BH+[RNH3+] [RNH,’] + [RNHJ
+
(1)
k’B[RNH21 [RNH3+] + [RNH,]
~’BH+
1 + K,/[H,O+]
k’BKa/
-
LH30+1
+ 1 + Ka/[H30f]
(2)
where kkH+,k’B, and Ka stand for the k’of the protonated amine, the k‘of the uncharged amine, and the acid dissociation constant, respectively. As uncharged species are retained much longer than charged species in RPLC, l-I4Nis expected to be retained longer than l-I5Nat a pH near the pKa where a difference in the extent of protonation exists between the two isotopic anilines due to the nitrogen isotope effect. N o difference in retention was seen between l-I4Nand I-I5N at pH 2.37 and 7.24, indicating that the nitrogen isotope effects on klBH+and k’B are negligible. At pH 4.24, Li4Ngave a k’of 5.869 compared to a k’of 5.81 1 with l - ” N (I4k’/l5k’ = 1.010). The nitrogen isotope effect on the dissociation of anilinium ion,
(2) Karger, B. L.; LePage, J. N.; Tanaka, N. ’High Performance Liquid Chromatography”; Horvith, C., Ed.; Academic Press: New York, 1980; pp 1 13-206. (3) 1-I4Nstands for ordinary aniline without any isotopic enrichment, unless its separation from 1-15N is specifically indicated. l-15N of 99% enrichment was purchased. (4) Cosmosil 5-CL8-P(Nakarai Chem.) was used. (5) Henry, R. A.; Byrne, S. H.; Hudson, D. R. J . Chromatogr. Sci. 1974, 12, 197-199. (6) The apparent pH of mobile phase was measured with a pH meter. The pH values were not included in the subsequent calculation of the isotope effect according to eq 3.
of l-I4Nand l-I5N,respectively. The nitrogen isotope effect on the dissociation of anilinium ion is comparable with the case of the oxygen isotope effect which gave I6Ka/l8Ka= 1.020 between benzoic acid-I6O2and benzoic acid-IsO,. The results may include the effect of the difference in the amount of methanol in solution, 5% in the present case compared to 20% in the case of the oxygen isotope effect. (The isotope effect on the dissociation of formic acid was shown to decrease with the increase in acetone content.*) Further study is needed to obtain these isotope effects without added organic solvents in the chromatographic mobile phase.’ Figure 1 shows the chromatograms of the isotopic mixture of aniline after 5 and 20 recycles on the 60-cm column (effective column lengths of 3 and 12 m). Base-line separation was achieved in the latter case, suggesting the possibility of facile enrichment of ISN isotope from the natural abundance of 0.37%. In this case the elution order is favorable for the recovery of I5N-enriched material, because the tail of the major peak of I4N compound does not decrease the isotopic purity of the enriched minor isotope, as might be the case with the oxygen isotope separation where the minor isotope eluted later.’ Although the separation factor obtainable by the present method based on the ionization control may be smaller than that in the conventional ion exchange process for the ammonium ion,‘O the wide availability of high-performance columns for RPLC can make this approach more attractive for the preparation of I5N-enriched compounds in laboratory. Together with the results with the oxygen isotope separation described earlier,’ the ionization control seems to be applicable for the isotopic separation of most weak acids and bases which undergo partial dissociation in aqueous media. Although most isotopic separations for oxygen and nitrogen have been carried out with simple compounds such as water, COz, NH3, and urea,” one can get isotopic enrichment for the compounds of interest by the present approach, instead of synthesizing them from enriched reagents. Although the use of silica-based packing materials will limit the applicability of ionization control to compounds with a pK, below 8, an organic polymer-based stationary phase will allow similar separations over a wider pH range. Such studies are currently in progress along with the measurements of heavy atom isotope effects in acid-base dissociation equilibria. Acknowledgment, We acknowledge Dr. J. W. Dolan for reviewing the manuscript. Registry No. l-I4N,62-53-3; l-I5N, 7022-92-6; ISN, 14390-96-6 (7) k’values were normalized by using the retention of hydroquinone in each mobile phase as a standard in order to eliminate the effect of the slight variation in methanol content in each mobile phase. k’values at pH 2.37 and 7.24 were used as k’BH+and .k’B, respectively, in eq 2. (8) Ellison, S. L. R.; Robinson, M. J. T. J Chem. SOC.,Chem. Commun. 1983,745-746. (9) Elution without added methanol was accompanied by either prolonged retention times or poor peak shape with currently available packing materials. Nitrogen isotope effects on the dissociation of protonated amines including phenylalanine and pyridine are in a range of 1.016-1.021 in water (Hermes, J. D.; Weiss, P. M.; Cleland, W.W.Biochemistry 1985,24, 2959-2967) and an unpublished observation by Kurz, J. L. See also: Ando, T.; Yamataka, H.; Wada, E. Isr. J . Chem., in press. (10) (a) Spedding, F. H.; Powell, J. E.; Svec, H. J. J . A m . Chem. SOC. 1955, 77, 6125-6132. (b) Ishimori, T. Bull. Chem. Soc. Jpn. 1960,33, 516-519. (c) Gupta, A. R.; Sarpal, S. K. J . Phys. Chem. 1967,71, 50G508. (d) Kotaka, M.; Shono, T.; Kakihana, H. Chem. Lett. 1978,315-318. (e) Trivelin, P. C.0.; Matsui, E.; Salati, E. Energ. Nucl. Agric. 1979,1, 1-13; Chem. Abstr. 1979,91, 10769~.(f) Hirschberg, K.; Krumbiegel, P.; Faust, H. Isoropenpraxis 1981,17, 178-182; Chem. Abstr. 1981,95, 57380q. (1 1) Bairamov, V. T.; Gorshkov, V. I.; Mal’tseva, V. V.; Belousova, M. E. Zh. Fir. Khim. 1984, 58, 1794-1795; Chem. Absrr. 1984, 101, 139203r.
